1
Correlation of Clinical and MRI Findings in Hips of 1
Elite Female Ballet Dancers 2
3
Victoria B. Duthon† MD, Caecilia Charbonnier* PhD, Frank C. Kolo
‡ MD, Nadia 4
Magnenat-Thalmann* PhD, Christophe D. Becker‡ MD, Cindy Bouvet
× MS, Elia 5
Coppens× MS, Pierre Hoffmeyer
† MD, Jacques Menetrey
† MD 6
7
†Department of Orthopaedic Surgery, University Hospital of Geneva, Geneva, 8
Switzerland 9
*MIRALab, University of Geneva, Geneva, Switzerland 10 ‡Department of Radiology, University Hospital of Geneva, Geneva, Switzerland, 11
×Medical University of Geneva, Switzerland 12
13
14
ABSTRACT 15
Purpose: To understand why professional female ballet dancers often complain of 16
inguinal pain and develop early hip osteoarthritis (OA). Goals were to examine 17
clinical and advanced imaging findings in the hips of dancers when compared to a 18
matched cohort of non-dancers; and to assess the femoral head translation in the 19
forward split position using Magnetic Resonance Imaging (MRI). 20
Methods: Twenty professional female ballet dancers and fourteen active healthy 21
female matched for age (control group) completed a questionnaire on hip pain, 22
underwent hip examination with impingement tests and measurement of passive hip 23
range of motion (ROM). All had a pelvic 1.5-T MRI in back-lying position to assess 24
femoroacetabular morphology and lesions. For the dancers, additional MR images 25
were acquired in split position to evaluate femoroacetabular congruency. 26
Results: 12 of 20 dancers complained of groin pain, only while dancing; controls 27
were asymptomatic. Dancers’ passive hip ROM was normal. No differences in α neck 28
angle acetabular depth, acetabular version and femoral neck anteversion were found 29
between dancers and controls. MRI of dancers while doing the splits showed a mean 30
femoral head subluxation of 2.05 mm. MRI of dancers’ hips showed labral tears, 31
cartilage thinning, and herniation pits, located in superior and postero-superior 32
positions. Lesions were the same for symptomatic and asymptomatic dancers. 33
Controls had proportionally the same amount of labral lesions, but in antero-superior 34
position. They also had 2 to 3 times less cartilage lesions and pits than dancers. 35
Conclusions: The results of our study are consistent with our hypothesis that 36
repetitive extreme movements can cause femoral head subluxations and 37
femoroacetabular abutments in female ballet dancers with normal hip morphology, 38
which could result in early OA. Pathological changes seen on the MRI were 39
symptomatic in less than two thirds of the dancers. 40
Level of Evidence: IV 41
42
2
INTRODUCTION 43
Ballet is a combination of sport and art. Professional ballet dancers use extreme hip 44
range of motion (ROM), often beyond conventional physiologic limits, to achieve 45
ideal ballet technique and aesthetic. Ballet movements and postures are based on the 46
“turnout” position: the lower extremity is externally rotated, from the hip to the foot. 47
Ideal turnout involves 55° to 70° of external rotation (ER) at the hip, 10° of ER at the 48
knee, up to 12° of tibial torsion, and abduction of the forefoot at the midtarsal joint.1 49
Most dancers do not achieve adequate turnout and compensate by increasing the 50
lumbar lordosis, "screwing the knees", and/or rolling in of the foot, leading to the 51
most common overuse injuries in dancing: lumbar stress fracture or 52
spondylolisthesis2, medial knee ligaments injuries
3, patellofemoral pain, and foot and 53
great toe problems4. There is a high prevalence and incidence of lower extremity and 54
back injuries, with soft tissue and overuse injuries predominating. Lifetime prevalence 55
estimates for injury in professional ballet dancers range between 40% and 84%.5 Hip 56
problems form only 10% of ballet injuries in most published series.6 However, many 57
dancers complain of groin pain while dancing, mainly during dynamic movements 58
implying extreme hip flexion and abduction combined with external rotation, such as 59
"grand plié", "grand battement à la seconde" or "grand développé à la seconde" 60
(Figure 1). Ballet dancers are known to present a higher risk of developing early hip 61
osteoarthritis (OA).7 Femoroacetabular impingement (FAI) has been shown to be a 62
risk factor for early OA.8,9
FAI is more commonly found in patients with abnormal 63
acetabulum (retroverted, coxa profunda, protrusio) or with an abnormally shaped 64
proximal femur (poor head-neck offset, posttraumatic deformities, slipped capital 65
femoral epiphysis, femoral retrotorsion, coxa vara, femoral head necrosis with 66
flattening). Two types of FAI have been described: the cam type FAI (femoral neck 67
abnormality) and the pincer type FAI (acetabular rim abnormality).10
68
Femoroacetabular abutment can also be found in patients with normal hip anatomy, 69
but performing extreme hip ROM (e.g., ballet dancers, yoga practitioners).11
This 70
dynamic and activity-related femoroacetabular abutment could lead to early hip OA, 71
but the exact mechanism has never been demonstrated. There is also the hypothesis 72
that the femoral head subluxates in extreme positions, leading to increased cartilage 73
stress and with time to OA. 74
The aims of the study were (1) to clinically evaluate professional female dancers’ hips 75
with measurement of the passive ROM, (2) to search for femoroacetabular lesions that 76
might correlate the groin pain described by many dancers, (3) to investigate if 77
femoroacetabular joint congruency is preserved while doing extreme movements such 78
as in split position (i.e., when the dancer sits down on the floor with one leg straight 79
forward and the other straight back, at right angles to the trunk) and (4) to correlate 80
clinical findings to Magnetic Resonance Imaging (MRI) examination. Our hypothesis 81
was that extreme movements may cause femoroacetabular abutments and femoral 82
head subluxations, which could result in early OA in dancers with normal hip 83
morphology. 84
3
85
Figure 1. Left: a ballet dancer in “grand plié” position. Right: “developpé à la seconde” 86 position. “Grand battement à la seconde” is the same movement but dynamic. Those 87 movements are a combination of flexion, abduction and external rotation of the hip. 88
89
90
MATERIALS AND METHODS 91
Study population and study design 92
Twenty female ballet dancers (mean age, 26 years; age range, 18-39 years) were 93
recruited and fourteen active healthy female matched for age as a control group (mean 94
age, 27 years; age range, 20-34 years). The dancers were either advanced students at 95
higher schools of dance or professional dancers. They all performed ballet and 96
contemporary dance. All had been dancing for more than 10 years and practiced for 97
more than 12 hours per week. Exclusion criteria for the volunteers of both groups 98
were hip injury, prior hip surgery, and any usual contraindication for MRI. The study 99
was approved by the local ethics committee and the volunteers gave written informed 100
consent. 101
102
Clinical evaluation 103
- Questionnaire 104
Participants had to complete a questionnaire (Table 1) made by the authors about the 105
presence (side, intensity) and localization of hip pain (“inguinal” for groin pain, 106
“lateral” for pain around the greater trochanter, “buttock” for posterior pain), activities 107
which triggered the pain, and evaluation of consecutive activity limitations (e.g., stairs 108
climbing, sitting). The questionnaire also asked dancers about the chronological 109
relation of the hip pain with their dancing activities. 110
- Physical examination 111
Dancers underwent a complete physical examination of the hip with measurement of 112
passive ROM in flexion/extension, abduction/adduction (back-lying with hip and knee 113
in extension) and internal/external rotation (back-lying with hip and knee flexed at 114
90°). While one physician was holding position of the lower limb, a second one 115
measured hip angles in those different positions with a hand held goniometer. Each 116
measure was done twice to verify the repeatability of the measure, and it was a 117
4
consensus exam. Care was taken to stabilize the pelvis during passive motion to 118
prevent overestimation of hip range of motion. The results were correlated to the 119
normal value for each movement according to Debrunner.12
120
Anterior impingement test was done for each volunteer, looking at elicited pain.11
The 121
test was done in supine position, with 90° to 120° of hip flexion and maximal 122
adduction and internal rotation. When groin pain was elicited by anterior 123
impingement test, the test was considered as positive. Posterior impingement test was 124
done in supine position with hyper-extension and external rotation. When buttock 125
pain was elicited by posterior impingement test, the test was considered as positive. 126
Internal snapping hip was tested by passively taking the hip from a flexed/externally 127
rotated position to an extended/internally rotated position. Other causes of groin pain 128
such as adductor tendinopathy, symphisitis or sports hernia were controlled. 129
130
Radiological evaluation 131
MRI of the pelvis in supine position was performed with a 1.5-T system (Avanto; 132
Siemens Medical Solutions, Erlangen, Germany) to assess femoroacetabular 133
morphology and lesions in dancers and controls. For the dancers, additional images 134
were acquired in split position to evaluate the hip joint in this extreme position 135
(Figure 2). 136
137
Figure 2. A ballet dancer in split position before MR imaging. 138 139
Two musculoskeletal radiologists performed a consensus reading with two 140
randomized patient’s orders and without any information about the clinical evaluation. 141
- Morphology 142
The morphology of both the femoral head and the acetabulum was measured on MRI 143
for each volunteer according to radiographic criteria: acetabular depth13
, acetabular 144
version14
and femoral α neck angle.15
The acetabular depth was evaluated according 145
to the method detailed by Pfirrmann et al.13
The depth was considered as positive and 146
normal if the center of the femoral head (O) was lateral to the line connecting the 147
anterior and posterior acetabular rim (Figure 3A). Measurement of the acetabular 148
version was based on the angle between the sagittal direction and lines drawn between 149
5
the anterior and posterior acetabular rim, at different heights (Figure 3B). The angle 150
was considered as positive (anteversion) when inclined medially to the sagittal plane 151
or negative (retroversion) when inclined laterally to the sagittal plane. The α neck 152
angle was measured in accordance with the method described by Notzli et al.15
in 153
eight positions (1: anterior, 2: anterosuperior, 3: superior, 4: posterosuperior, 5: 154
posterior, 6: posteroinferior, 7: inferior, 8: anteroinferior) around the femoral neck 155
(Figure 3C,D). Deviation from the normal geometry was associated with α angles 156
larger than 55°. 157
158
Figure 3. A) Definition of the acetabular depth on a transverse oblique MR image obtained 159 through the center of the femoral neck according to Pfirmann et al.
13 B) Computation of the 160
acetabular version: roof edge (RE) and equatorial edge (EE) are lines drawn between the 161 anterior and posterior acetabular edges, defining the orientation of the acetabular opening 162 proximally and at the maximum diameter of the femoral head respectively (arrows). C) 163 Definition of the α angle on a radial MR image according to Notzli et al.
15 illustrating a cam 164
type morphology ( = 85°). D) Acetabulum divided into 8 sectors (1:anterior, 165 2:anterosuperior, 3:superior, 4:posterosuperior, 5:posterior, 6:posteroinferior, 7:inferior, 166 8:anteroinferior) where α angles were measured. The sectors were also used to document 167 locations of labrum and cartilage abnormalities. 168
169
In addition to those cam/pincer indicators, 2 measurements were performed to verify 170
the presence of any other morphological features: femoral neck-shaft angle (Figure 171
4A) and neck anteversion defined in the axial plane by superimposing MR images 172
taken at different heights (Figure 4B). 173
- Pathology 174
For each subject, labrum and cartilage abnormalities were assessed qualitatively at 175
eight positions, as depicted in Figure 3D. Acetabular cartilage was considered as 176
normal (grade 0), hyperintense (grade 1), thinning (grade 2), tear (grade 3), flap 177
6
(grade 4), or hyposignal (grade 5), and extent of cartilage damage was documented. 178
The acetabular labrum was considered as normal (grade 0), degenerated (abnormal 179
signal intensity, grade 1), torn (abnormal linear intensity extending to the labral 180
surface, grade 2), as ossification of the labrum (continuity of the labrum with 181
acetabular bone marrow, grade 3), or as a separated ossicle (os acetabuli, grade 4). 182
The presence of subchondral acetabular or femoral bony abnormalities (e.g., edema, 183
cysts) and the presence of a herniation pit (a round cystic lesion at the anterior aspect 184
of the femoral neck) were also reported.10
185
186
Figure 4. A) Computation of the neck-shaft angle on a frontal MR image. B) Definition of the 187 femoral neck anteversion, defined by the angle formed by the line O-O’ connecting the center 188 of the femoral head (O) and the center of the femoral neck (O’) at its narrowest point, and the 189 line MC-LC connecting the medial condyle (MC) and the lateral condyle (LC). This angle is 190 calculated in the axial plane by superimposing MR images taken at different heights. 191
192
- Femoroacetabular congruency 193
Dancers underwent MRI in split position to evaluate congruency of the hip joint in 194
such extreme position. Femoral head subluxation was evaluated by the method 195
described by Gilles et al.16
The method is based on a surface registration technique 196
and required the following procedure (Figure 5): first, a virtual 3D model of the hip 197
joint is reconstructed using the MR images in supine position, thanks to a validated 198
segmentation software.17,18
Thus, for each dancer, patient-specific 3D models of the 199
pelvis and femur were obtained. Second, the hip joint center (HJC) position is 200
estimated in this reference neutral posture. Third, the 3D bony models are registered 201
to extract joint poses from MR images in split position. Finally, femoroacetabular 202
translations are measured with reference to the previously estimated HJC. The 203
interested reader can refer to the article for more details.16
204
205
Statistical analysis 206
A statistical analysis was conducted to compare dancers’ hip morphology to controls. 207
Alpha neck angles measured in eight positions were compared. Acetabular depth, 208
acetabular version, femoral neck anteversion and femoral neck-shaft angle were also 209
compared between the two groups. For all comparisons, p-values were calculated with 210
the Mann-Whitney U test and were considered as statistically significant if < 0.05. 211
7
212
Figure 5. A virtual 3D model of the hip joint is reconstructed using the MR images in supine 213 position (reference neutral posture, left). The HJC is estimated in this position. The 3D 214 models are registered to extract joint poses from MR images in split position (right). 215 Femoroacetabular translations are then measured with reference to the previously estimated 216 HJC. 217
218
RESULTS 219
Clinical results 220
- Questionnaire 221
The results of the questionnaire for each dancer are listed in Table 2. Twelve out of 20 222
dancers complained of hip pain: 4 bilaterally, 7 on the right hip and 1 on the left hip. 223
Pain was inguinal and occurred only while dancing, mainly at the end of the ROM of 224
specific dancing movements, such as "grand battement à la seconde", "grand plié" and 225
"développé à la seconde". It is worth mentioning that all these movements imply 226
extreme abduction and flexion combined with external rotation of the hip. Dancers 227
had no pain and no limitations during daily living (such as while sitting, squatting, 228
crossing legs, sleeping, climbing stairs, etc.). Control group was asymptomatic. 229
- Physical examination 230
The results of the dancers’ passive hip ROM are listed in Table 3, with the normal 231
ROM taken as reference.12
The dancers had normal passive hip ROM, with a trend to 232
increased abduction and external rotation (50% were over the normal range), and to 233
decreased internal rotation (30% were below the normal range). Pain could be 234
reproduced by the anterior impingement test for 7 out of 20 dancers (bilateral for 2 235
dancers). Posterior impingement test was positive for 3 dancers (unilateral). Internal 236
snapping hip was present in 2 dancers (unilateral), but not painful. Other causes of 237
groin pain such as adductor tendinopathy, symphisitis or sports hernia were not found. 238
Control group was by definition asymptomatic, and anterior and posterior 239
impingement tests were not painful. 240
241
8
Radiological results 242
Radiological analysis was performed on 39 dancer’s hips (20 dancers, one hip 243
excluded due to a technical problem during MRI scan) and on 28 controls’ hips (14 244
controls). 245
- Morphology 246
Dancers’ femoral neck anteversion, acetabular version and acetabular depth were 247
normal and comparable to controls. Femoral neck-shaft angle was lower in dancers 248
than in controls. This result was statistically significant (Table 4). Alpha neck angles 249
in eight positions were analyzed and compared between both groups. Mean α neck 250
angles in those positions were normal in dancers and controls. Cam morphology was 251
found in only one dancer (maximal values), none in the control group (Table 5). 252
Dancers had lower α neck angles than controls in anterior, superior, postero-superior, 253
postero-inferior, inferior and antero-inferior positions, and this was statistically 254
significant. 255
- Pathology 256
MRI of the dancers’ hips revealed 3 types of lesions: 1) degenerative labral lesions, 2) 257
acetabular cartilage thinning with subchondral cysts, and 3) herniation pits. 258
Degenerative labral lesions were found for both dancers and controls, in superior 259
position. However, cartilage lesions were twice more frequent and more severe in 260
dancers than in controls: 75% of dancers had cartilage tear in superior position (some 261
in postero-superior position), and 28% of controls had cartilage thinning (i.e., less 262
severe than dancers) in antero-superior position. Herniation pits were more than twice 263
more frequent in dancers than in controls (Table 6). 264
- Femoroacetabular congruency 265
Dancers while doing the splits in the MRI tube allowed us to assess static 266
femoroacetabular congruency in this extreme position. This analysis showed a mean 267
femoral head subluxation of 2.05 mm (range 0.63-3.56 mm), according to the method 268
described by Gilles et al.16
We did not observe any privileged direction of 269
femoroacetabular translations. 270
271
Correlation between clinical and radiological results 272
Correlation of clinical and MRI findings led us to classify dancers in 4 groups (Table 273
7): 1) pain with lesions on MRI, 2) pain without lesions on MRI, 3) no pain but 274
lesions on MRI, 4) no pain and no lesion on MRI. 90% of dancers presented labral 275
and/or cartilaginous lesions on MRI; however, only 61% of them were symptomatic. 276
No criteria in the data was found to explain why some dancers having 277
femoroacetabular lesions were painful, while others with the same lesions were not; 278
indeed, lesions on MRI were the same for symptomatic and asymptomatic dancers. 279
280
DISCUSSION 281
This study resulted in several interesting findings: dancers had normal passive hip 282
ROM and bony morphology; 60% of them complained of groin pain while dancing; 283
most of them had labral and/or cartilaginous lesions and showed a femoral head 284
9
subluxation during extreme movements; no correlation between clinical and 285
radiological findings could be done. 286
In this study, dancers’ hip ROM was normal and we noted a tendency to increased 287
external rotation and abduction (50% were above the normal range), and to decreased 288
internal rotation (30% were below the normal range). This can be explained by 289
constant external rotation of the lower limb in ballet, also called the “turnout” 290
position, as already described in previous studies.19
Dancers train from childhood to 291
achieve ideal turnout position. Hamilton et al.20
showed that dancers who trained for 6 292
hours a week or more at 11-14 years old had significantly less femoral torsion, and 293
then had greater passive external rotation, but this had no influence on the execution 294
of turnout. In our study, the dancers did not present less femoral antetorsion than 295
controls. Therefore, we believe that the trend of increased external rotation and their 296
capacity to achieve turnout and extreme movements may be due to soft tissue 297
adaptation (ligament flexibility and muscle strength) achieved by training during 298
years.21
In addition, it is important to note that extreme movements such as 299
“développé à la seconde” or “grand battement à la seconde” (see Figure 1) can be 300
achieved thanks to a combination of abduction and external rotation of the hip. Pure 301
abduction would be limited to normal range of motion (30-50°), but thanks to the 302
combination of abduction, flexion and external rotation, dancers can reach an extreme 303
position during “grand battement à la seconde”. 304
Interestingly, it is precisely during this kind of movement that symptomatic dancers 305
complained of groin pain. This let us think that an abnormal femoroacetabular contact 306
may occur during these movements and induce pain in the joint. A French article 307
published in 1979 described hip pain associated with structural abnormalities of the 308
proximal femoral neck in athletes participating in hockey, football, soccer, rugby, 309
martial arts, and tennis22
. In the present study, MRI allowed us to assess if dancers 310
had morphological abnormalities, such as FAI, explaining their hip lesions and pain. 311
As opposed to what stated Demarais22
, the dancers in our study had pain and/or 312
lesions in spite of normal hip morphology, except for one dancer with a cam FAI. 313
Dancers’ labral and acetabular cartilaginous lesions were the same kind of lesions as 314
those found in patients with FAI.20
However, the lesions were located in the 315
superior/postero-superior position of the acetabular rim, contrary to usual 316
anterior/antero-superior lesions found in cam or pincer FAI.23,24
Such lesions may be 317
explained by repetitive extreme movements combining abduction and external 318
rotation, such as “grand battement à la seconde”. Indeed, Safran et al.25
confirmed that 319
the greatest strain laterally were at 90° of flexion with abduction and external or 320
neutral rotation. Therefore, the femoral neck may come in abutment with the 321
acetabular rim during dancing movements, leading to a superior/posterosuperior 322
dance-related femoroacetabular abutment. Moreover, Safran et al.25
showed that when 323
the hip was externally rotated, the posterior labrum had significantly increased strain. 324
Yet dancers have to permanently keep hip external rotation while dancing (i.e., the 325
“turnout” position), this could hence explain why their lesions were found in superior 326
and postero-superior positions. 327
Another interesting finding in this study was the mean 2 mm femoral head 328
subluxation in split position. Dynamic analysis of the same dancers with motion 329
capture showed us that this femoral head subluxation is even greater during dynamic 330
extreme movements such as “grand battement à la seconde”, and seems to be a 331
consequence of femoroacetabular abutment.26
Femoral head subluxation may also be 332
due to capsular laxity leading to hip micro-instability.27
Indeed McCormack et al. 333
10
showed that the prevalence of hypermobility and “benign joint hypermobility 334
syndrome” is significantly higher in dancers than controls, having an important 335
negative influence and leading to arthralgia.28
Then, repetitive subluxations could be a 336
cause of pain, acetabular cartilage lesions as seen in this study, and consequent early 337
OA. But early hip OA may also simply be due to overuse with repetitive motion and 338
recurrent impact, as found in other professional sports.29
339
We did also not find correlation between radiological and clinical findings. Dancers 340
groin pain can neither be explained by the type of lesions nor by the ROM and bone 341
morphology, as there was no difference between the 4 groups. In addition, MRI of 342
controls showed that labral and acetabular lesions could be found in asymptomatic 343
patients. Pain can also be caused by muscular and ligamentous structures around the 344
hip joint. Bedi et al.30
showed that hip pain, in the absence of OA, may be due to a 345
complex combination of mechanical stresses, both dynamic and static, leading to 346
compensatory dysfunction of the periarticular musculature. Due to the constant 347
“turnout” position, dancers have increased external hip rotation and powerful external 348
rotator muscles, and this dysbalance between internal and external rotators could be a 349
cause of hip pain and could be prevented by a more balanced muscle strengthening 350
and stretching. 351
In summary, prevention could be done by limiting these extreme movements implying 352
femoroacetabular abutment and subluxation and leading with time to early OA. 353
Dancers should be aware of the fine line between maximizing the range and variety of 354
movement versus exceeding their physical limits, thus risking injuries. 355
Several study limitations need to be stated: 1) the small number of participants (20 356
dancers and 14 controls), a female study only31
; 2) the questionnaire on hip pain that 357
was made by the authors and was not referenced in the literature; 3) the radiological 358
analysis that was based on native hip MRI (reliability of the findings estimated at 359
65%) and not MR arthrography. 360
361
CONCLUSIONS 362
The results of our study are consistent with our hypothesis that repetitive extreme 363
movements can cause femoral head subluxations and femoroacetabular abutments in 364
female ballet dancers with normal hip morphology, which could result in early OA. 365
Pathological changes seen on the MRI were symptomatic in less than two thirds of the 366
dancers.. 367
368
11
REFERENCES 369
1. Stephens RE. The etiology of injuries in ballet. In: Dance medicine- a 370
comprehensive guide. Chicago: Pluribus, 1987. 371
2. Micheli LJ. Back injuries in dancers. Clin Sports Med 1983;2:473-484. 372
3. Miller EH, Schneider HJ, Bronson JL, McLain D. A new consideration in athletic 373
injuries. The classical ballet dancer. Clin Orthop Relat Res 1975;111:181-191. 374
4. Hardaker WT Jr. Foot and ankle injuries in classical ballet dancers. Orthop Clin 375
North Am 1989;20:621-627 376
5. Hincapie CA, Morton EJ, Cassidy JD. Musculoskeletal injuries and pain in 377
dancers: a systematic review. Arch Phys Med Rehabil 2008;89:1819-1829. 378
6. Reid DC. Prevention of hip and knee injuries in ballet dancers. Sports Med 379
1988;6:295-307. 380
7. Binningsley D. Tear of the acetabular labrum in an elite athlete. Br J Sports Med 381
2003;37:84-88. 382
8. Ganz R, Parvizi J, Beck M, Leunig M, Nötzli H, Siebenrock KA. 383
Femoroacetabular impingement: a cause for osteoarthritis of the hip. Clin Orthop 384
Rel Res 2003;417:112-120. 385
9. Beck M, Kalhor M, Leunig M, Ganz R. Hip morphology influences the pattern of 386
damage to the acetabular cartilage: femoroacetabular impingement as a cause of 387
early osteoarthritis of the hip. J Bone Joint Surg Br 2005;87:1012-1018. 388
10. Leunig M, Beck M, Kalhor M, Kim YJ, Werlen S, Ganz R. Fibrocystic changes 389
at anterosuperior femoral neck: prevalence in hips with femoroacetabular 390
impingement. Radiology 2005;236:237-246. 391
11. Sierra RJ, Trousdale RT, Ganz R, Leunig M. Hip disease in the young, active 392
patient: evaluation and nonarthroplasty surgical options. J Am Acad Orthop Surg 393
2008;16:689-703. 394
12. Debrunner HU. Gelenkmessung (Neutral-O-Methode), Längenmessung, 395
Umfangmessung. Bern: Bulletin der Schweizerischen Arbeitsgemeinschaft für 396
Osteosynthesefragen, 1971 397
13. Pfirrmann CW, Mengiardi B, Dora C, Kalberer F, Zanetti M, Hodler J. Cam and 398
pincer femoroacetabular impingement: characteristic MR arthrographic findings 399
in 50 patients. Radiology 2006;240:778-785. 400
14. Reynolds D, Lucas J, Klaue K. Retroversion of the acetabulum. A cause of hip 401
pain. J Bone Joint Surg Br 1999;81:281-288. 402
15. Notzli HP, Wyss TF, Stoecklin CH, Schmid MR, Treiber K, Hodler J. The 403
contour of the femoral head-neck junction as a predictor for the risk of anterior 404
impingement. J Bone Joint Surg Br 2002;84:556-560. 405
16. Gilles B, Kolo FC, Magnenat-Thalmann N, Becker CD, Duc SR, Menetrey J, 406
Hoffmeyer P. MRI-based assessment of hip joint translations. J Biomech 407
2009;42:1201-1205. 408
12
17. Gilles B, Moccozet L, Magnenat-Thalmann N. Anatomical modelling of the 409
musculoskeletal system from MRI. Med Image Comput Comput Assist Interv 410
2006;9:289-296. 411
18. Schmid J, Magnenat-Thalmann N. MRI bone segmentation using deformable 412
models and shape priors. Heidelberg: Springer Berlin, 2008. 413
19. Khan K, Roberts P, Nattrass C, Bennell K, Mayes S, Way S, Brown J, 414
McMeeken J, Wark J. Hip and ankle range of motion in elite classical ballet 415
dancers and controls. Clin J Sport Med 1997;7:174-179. 416
20. Hamilton D, Aronsen P, Loken JH, Berg IM, Skotheim R, Hopper D, Clarke A, 417
Briffa NK. Dance training intensity at 11-14 years is associated with femoral 418
torsion in classical ballet dancers. Br J Sports Med 2006;40:299-303. 419
21. Bennel KL, Khan KM, Matthews BL, Singleton C. Changes in hip and ankle 420
range of motion and hip muscle strength in 8-11 year old novice female ballet 421
dancers and controls: a 12 month follow up study. Br J Sports Med 2001;35:54-422
59. 423
22. Demarais Y, Lequesne M. La hanche du sportif. Gaz Med France 86 :2969-2972. 424
1979. 425
23. Beaulé PE, Zaragoza E, Motamedi K, Copelan N, Dorey FJ. Three-dimensional 426
computed tomography of the hip in the assessment of femoroacetabular 427
impingement. J Orthop Res 2005;1286-1292 428
24. Tannast M, Goricki D, Beck M, Murphy SB, Siebenrock KA. Hip damage occurs 429
at the zone of femoroacetabular impingement. Clin Orthop Relat Res 430
2008;466:273-280 431
25. Safran MR, Giordano G, Lindsey DP, Gold GE, Rosenberg J, Zaffagnini S, Giori 432
NJ. Strains across the acetabular labrum during hip motion: a cadaveric model. 433
Am J Sports Med 2011;39:92S-102S. 434
26. Charbonnier C, Kolo FC, Duthon VB, Magnenat-Thalmann N, Becker CD, 435
Hoffmeyer P, Menetrey J. Assessment of congruence and impingement of the hip 436
joint in professional ballet dancers: a motion capture study. Am J Sports Med 437
2011;39:557-566. 438
27. Smith MV, Sekiya JK. Hip instability. Sports Med Arthrosc 2010;18:108-112. 439
28. McCormack M, Briggs J, Hakim A, Grahame R. Joint laxity and the benign joint 440
hypermobility syndrome in student and professional ballet dancers. J Rheumatol 441
2004;31:173-178. 442
29. Lindberg H, Roos H, Gärdsell P. Prevalence of coxarthrosis in former soccer 443
players. 286 players compared with matched controls. Acta Orthop Scand 444
1993;64:165-167. 445
30. Bedi A, Dolan M, Leunig M, Kelly BT. Static and dynamic mechanical causes of 446
hip pain. Arthroscopy 2011;27:235-251. 447
31. Nilsson C, Leanderson J, Wykman A, Strender LE. The injury panorama in a 448
Swedish professional ballet company. Knee Surg Sports Traumatol Arthrosc 449
2001;9:242-246. 450
451
13
Table 1. Questionnaire on hip pain. 452
Do you have hip pain? yes / no
If you do, which side? left / right
Which intensity? 0-1-2-3-4-5-6-7-8-9-10
Since when?
For how long are you a professional dancer?
What are your past health problems?
How is your general health?
How long can you walk without pain? _______min.
Do you have pain when you climb up stairs? yes / no
If you do, where do you feel the pain? inguinal / lateral / buttock
Which side? right / left
Do you have pain when you go down stairs? yes / no
If you do, where do you feel the pain? inguinal / lateral / buttock
Which side? right / left
Do you have pain when you stay squatted? yes / no
If you do, where do you feel the pain? inguinal / lateral / buttock
Which side? right / left
Do you feel pain when you cross legs? yes / no
If you do, where do you feel the pain? inguinal / lateral / buttock
Which side? right / left
Do you have night pain? yes / no
Which other movement/position is painful?
453
454
14
Table 2. Dancers’ answers to the questionnaire on hip pain. 455
Hip pain Side of hip pain Localization of pain Occurrence
Dancers Yes No Right Left Inguinal Lateral Buttock Dance Daily living
1 x x x x
2 x x x x
3 x
4 x
5 x
6 x x x x
7 x x x x
8 x x x x
9 x
10 x
11 x
12 x
13 x x x x x
14 x x x x x
15 x x x x
16 x x x x (right) x (left) x
17 x x x x
18 x x x x
19 x x x x x
20 x
456
15
Table 3. Dancers’ passive hip range of motion measured in degree by clinical 457 examination. Dancers tend to have increased abduction and external rotation, and 458
decreased internal rotation when compared to normal range. 459
Hip ROM Min Mean SD* Max Normal range
Flexion 115 133 10 150 120-140
Extension 10 19 4 25 10-20
Abduction 30 61 20 100 30-50
Adduction 10 25 13 45 20-30
Internal rotation 5 33 11 50 30-45
External rotation 30 56 13 80 40-50
*SD: standard deviation.460
16
Table 4. Dancers’ and controls’ hip morphology measured on MRI. 461
Hip morphology Dancers (n=39)* Controls (n=28)* p-value†
Mean SD‡ Mean SD
‡
Femoral neck-shaft angle (°) 132.5 4.6 135.4 3.4 0.003
Femoral neck anteversion (°) 12.2 6 13.9 7.7 0.386
Acetabular depth (mm) 7.8 1.6 8.8 2.2 0.065
Acetabular version (°) 6.7 5.3 6 5 0.172
*Data are the number of hips 462 ‡SD: standard deviation. 463
†P-value obtained with use of Mann-Whitney U test. P-value < 0.05 means that there is a statistically 464
significant difference between dancers and controls.465
17
Table 5. Alpha neck angles (degree) measured on MRI for dancers and controls. The eight positions are illustrated in Figure 3D. 466
Dancers (n=39)* Controls (n=28)* p-value†
Position Min Mean SD‡ Max Min Mean SD
‡ Max
Anterior 36.5 45.6 5.5 66.3 39 47.5 4 55.1 0.022
Antero-superior 34.9 47.4 7.4 76 35 46 5 54.6 0.550
Superior 32.1 40.9 5.1 54.6 37.8 46.6 4.5 55.3 0.001
Postero-superior 31.2 37.8 3.7 44.1 33.1 43.1 6.7 59.7 0.001
Posterior 30.3 39.2 4.3 48.4 33.8 40.3 4.8 49.6 0.538
Postero-inferior 28.8 38 3.7 48.3 36.7 48.7 7 64.2 0.001
Inferior 32.5 39.9 3.7 48.2 42 51.2 6.3 62.9 0.001
Antero-inferior 32 40.6 3.3 46 35.9 44.7 5.4 57.3 0.002
*Data are the number of hips
467 ‡SD: standard deviation. 468
†P-value obtained with use of Mann-Whitney U test. P-value < 0.05 means that there is a statistically significant difference between dancers and controls.469
18
Table 6. Percentage of dancers and controls having labral lesions, acetabular 470 cartilage lesions, and herniation pits. For labral lesions, intensity 1 means 471
degeneration (abnormal signal intensity) and intensity 2 means tear (abnormal linear 472
signal intensity extending to the labral surface). For acetabular cartilage lesions, 473
intensity 1 means hyperintensity, intensity 2 means thinning and intensity 3 means 474
tear. 475
Dancers (n=39)* Controls (n=28)* p-value†
Labral lesions
Incidence 33/39 (85%) 24/28 (85%) 1
Mean intensity 1.73 1.76 0.745
Acetabular
cartilage lesions
Incidence 29/39 (75%) 8/28 (28%) 0.037
Mean intensity 2.67 1.83 0.007
Pits Incidence 23/39 (60%) 6/28 (21%) 0.038
*Data are the numbers of hips
476 †P-value obtained with use of Mann-Whitney U test. P-value < 0.05 means that there is a statistically 477
significant difference between dancers and controls. 478 479
19
Table 7. Repartition of the 20 dancers according to the presence of pain and/or 480
lesions on MRI. 481
Lesions on MRI No lesions on MRI
Pain 11 1
No pain 7 1
482